Path planning method and system for autonomous mobile robots based on cost map layers and recovery levels

The path planning method for AMRs simplifies and stabilizes path planning by using a cost map with activated layers to reflect zone-specific characteristics, reducing unnecessary path changes and improving efficiency in multi-robot environments.

KR102991547B1Active Publication Date: 2026-07-15SYSWIN MOBILITY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
KR1020260095694
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2026-05-27
Publication Date
2026-07-15
Estimated Expiration
2046-05-27

AI Technical Summary

Technical Problem

Existing path planning methods for autonomous mobile robots (AMRs) in industrial environments are complex and unstable due to varying driving zones and dynamic obstacles, leading to unnecessary detours, lane deviations, and path recalculation issues, especially when multiple robots operate simultaneously.

Method used

A path planning method for AMRs that determines driving area types and activates specific layers of a cost map, such as peer robot filter, lane constraint, and movement trajectory layers, forming a merged cost map to search for reference or alternative paths based on zone-specific characteristics, suppressing unnecessary path changes and re-entry.

Benefits of technology

This approach simplifies path planning by reducing complexity and maintaining stability, suppressing unnecessary path changes, and improving operational efficiency in multi-robot environments by reflecting zone-specific driving characteristics and avoiding lane deviations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 112026064172305-PAT00001_ABST
    Figure 112026064172305-PAT00001_ABST
Patent Text Reader

Abstract

The present invention relates to a path planning method for an autonomous mobile robot, comprising: a step of determining a driving area type of the autonomous mobile robot; a step of determining, based on the driving area type, the active state of at least one of a peer robot filter layer, a lane constraint layer, and a movement trajectory layer among a plurality of layers constituting a cost map; a step of forming a merged cost map by reflecting the active layer among the peer robot filter layer, the lane constraint layer, and the movement trajectory layer; a step of searching for a reference path from a starting posture to a target posture based on the cell cost included in the merged cost map; and a step of determining a recovery level corresponding to the driving area type when it is determined that the reference path is not generated or that obstacle avoidance is required for the reference path. The present invention relates to a path planning method for an autonomous mobile robot, comprising the step of determining whether to search for an alternative path or determining the search range of an alternative path according to the recovery level, wherein the recovery level is determined as any one of a first recovery level in which no search for an alternative path is performed, a second recovery level in which an alternative path is searched within an area limited by the merged cost map, and a third recovery level in which an alternative path is searched within a search range wider than the second recovery level.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Method for determining movement path to avoid collision with other unmanned vehicle

    KR1020240062983A

  • Path planning method and navigation method and mobile machine using the same

    US20250390099A1